When you visit any website, it may store or retrieve information on your browser, mostly in the form of cookies. This information might be about you, your preferences or your device and is mainly used to make the site work as you expect it to. The information does not directly identify you but can give you a more personalized web experience. Because we respect your right to privacy, you can choose not to allow some types of cookies. Click on the different category headings to learn more and change our default settings. However, blocking some types of cookies may impact your experience of the site. You can find more information, including a detailed cookie explanation, on our Cookie Policy.
On World Lung Cancer Month, it’s crucial to highlight advancements in the fight against this lethal disease.
Among all cancers, lung cancer is one of the most prevalent and deadly types globally. Early detection remains an essential milestone for patients suffering from the disease, since the primary cause of poor prognosis is late diagnosis. Indeed, about 75% of patients are diagnosed at advanced stages (stage III–IV), with a 5-year survival rate of less than 10% for stage IV. Consequently, strategies focused on identifying those patients at the stages that are curable (stages 0-II) are key to reducing lung cancer mortality.
This was the aim of a research led by investigators from the Guangzhou Medical University of China. The researchers developed PulmoSeek Plus, a composite model resulting from the combination of two models previously developed by the investigators: CIBM (based on clinical and imaging features) and PulmoSeek (based on circulating tumor DNA methylation biomarkers). Pulmoseek Plus integrates the predictive powers of clinical, imaging (low-dose computerized tomography) and cell-free DNA (cfDNA) methylation biomarkers using machine learning algorithms to classify malignant and benign pulmonary nodules.
Figure: Decision curve analysis plot for PulmoSeek Plus, PulmoSeek, and CIBM models: The plot shows the net benefit (y-axis) across a range of risk thresholds (x-axis) of three models compared with intervention in all participants (all) or no intervention (none) in the training set (A) and validation set (B). CIBM=clinical and imaging biomarkers.
The PulmoSeek Plus model was superior to the PulmoSeek and CIBM models. Decision curve analysis showed a higher net benefit in identifying malignant pulmonary nodules, especially when the threshold probability was above 0.19 in the training set and 0.04 in the validation set. At a risk threshold of 0.54, PulmoSeek Plus correctly identified about 83 out of 100 individuals with malignant lung cancer, outperforming the other models. This indicates its effectiveness as a clinical tool for lung cancer diagnosis.
This innovative tool was trained and validated in a retrospective study involving 24 hospitals across 20 cities in China that included a total of 1380 participants with a single pulmonary nodule sized 5–30 mm. The results of this study, recently published in The Lancet Digital Health journal, revealed that PulmoSeek Plus displayed a high sensitivity of 0.98 (0.97–0.99) at a fixed specificity of 0.50 for ruling out lung cancer. Importantly, when the authors compared the performance of this model with that offered by CIBM and PulmoSeek, PulmoSeek Plus resulted in a better discrimination capacity, with an increase of 0.05 in the area under the curve (AUC; PulmoSeek Plus vs CIBM, 95% CI 0.022–0.087, p = 0.001; and PulmoSeek Plus vs PulmoSeek, 0.018–0.083, p = 0.002). Furthermore, benefits were maintained in early-stage lung cancer (stages 0 and I) and in patients with small nodules (5–10 mm nodules in size), often challenging in clinical practice.
As evidenced, PulmoSeek Plus has the potential to be an effective tool for the early clinical evaluation and treatment of lung cancer, for which non-invasively obtained blood samples and CT scans would be the only prerequisites. Being able to identify approximately 83 of 100 people with lung cancer (see see Figure below), this model underscores the utility of combining powerful methodologies, such as artificial intelligence, imaging scans and novel biomarkers, with conventional clinical tools for the early identification of lung cancer.
Read the full article here.
Accurate classification of pulmonary nodules by a combined model of clinical, imaging, and cell-free DNA methylation biomarkers: a model development and external validation study. Jianxing He, Bo Wang, Jinsheng Tao, Qin Liu et al. (2023) Lancet Digit Health 2023; 5: e647–56